Glioblastomas (GBMs) are genomically well characterized, yet heterogeneous, and exhibit profound resistance to all existing treatment strategies. The most effective therapeutics are radiation therapy (RT) and the alkylating agent temozolomide (TMZ), but progression typically occurs within months after initiating these treatments. The mechanisms underlying this profound resistance remain unknown, but genetic heterogeneity is likely a major contributor, as has been shown in other cancers. Unfortunately, little is known about how GBM genomes evolve with treatment. This information would be useful to guide development of strategies to avoid the development of resistance and to identify optimal therapeutic approaches in the recurrent setting. We hypothesize that somatic genetic profiles of GBMs that recur after treatment with RT and TMZ differ substantially from pre-treatment GBMs, and that the differences point to mechanisms by which GBMs resist these treatments. To test this, we propose to identify and functionally validate recurrent genetic changes associated with resistance using innovative genomic analysis tools and patient derived model systems. Our collaborative consortium has collected an unprecedented number of paired pre- and post-treatment human tumors (>200). We have also created more than 100 patient derived GBM models that will be treated to test for the emergence of recurrent resistance drivers. Preliminary data from both patient samples and models indicate substantial tumor evolution occurs during treatment and identify TP53, CHEK2 and other rational targets as candidate mediators of resistance. Collective analysis of the data will be used to address two Aims. In Aim 1, we will test the hypothesis that treatment with radiation and temozolomide leads to consistent genetic changes in human tumors using whole exome sequencing of paired pre- and post-treatment tumor samples to determine large-scale changes in population structures and single cell sequencing to evaluate the effects of these treatments on microheterogeneity. In Aim 2, we will test the hypothesis that genetic changes identified in post-treatment GBMs functionally contribute to RT and TMZ resistance in GBM using patient derived cell lines (PDCL) and patient derived xenografts (PDX). We will determine the effects of radiation and temozolomide on these models and their genomic hierarchies using deep sequencing and test the effects of candidate drivers of resistance both in vitro and in vivo. We will determine whether resistant clones exist prior to treatment or are stochastically induced using an innovative single cell barcoding approach to determine whether the evolution of clonal substructures is consistent across replicate experiments. These studies will create a comprehensive understanding of genetic evolution during standard-of-care therapy for GBM. They will inform diagnostic approaches for assignment of targeted therapeutics in the recurrent setting and identify genetic changes driving resistance. Therapeutic targeting of these novel resistance drivers could represent a rational approach to substantially improve our existing standard of care for GBM patients.